In [23]: # quant_trading.ipynb
Abstract
Quantitative trading firms pay the highest cash compensation in the ML labour market. A senior ML researcher at a top firm in a good performance year earns total compensation $950,000 to $1,900,000+, comprising base salary $250,000 to $400,000 and performance-tied cash bonus $700,000 to $1,500,000+. The bonus is tied to firm-wide P&L and can drop to substantially lower levels in bad years. There is no equity grant; the long-run upside is the partner track, which generates substantially higher compensation but requires 5 to 10 years of sustained P&L contribution to reach [1].
1 Bands from Bloomberg, Wall Street Journal, and Financial Times reporting on quant compensation; Levels.fyi Citadel, Levels.fyi Two Sigma, and Levels.fyi Jane Street, May 2026.
table qt-1 : compensation by firm tier
| Firm tier | Base | Good-year bonus | Total range |
|---|---|---|---|
| Top tier (Renaissance, Two Sigma, Citadel, Jane Street, DE Shaw)PhD typical from a narrow university set | $250k - $400k | $700k - $1.5M+ | $950k - $1.9M+ |
| Second tier (Millennium, Point72, Balyasny, BlueCrest, Hudson River, Jump)Hedge fund and market-maker tier | $200k - $320k | $300k - $700k | $500k - $1.0M |
| Bank systematic trading desks (Goldman, Morgan Stanley, JP Morgan)Lower bonus volatility, more banking-culture norms | $180k - $260k | $150k - $400k | $330k - $660k |
| Smaller systematic fundsWider variance; firm-survival risk | $170k - $240k | $100k - $300k | $270k - $540k |
section qt-2 : narrow pool dynamics
The structural cause of quant compensation is talent compression. Top quant trading firms hire from a narrow university set (Carnegie Mellon, MIT, Princeton, Cornell, Stanford, Chicago, NYU Courant, plus a handful of European programs at Cambridge, Oxford, ETH, EPFL) and almost exclusively from PhD or post-doc programs with research backgrounds in mathematical statistics, optimisation, stochastic calculus, or high-energy physics. The annual global supply of qualified candidates from this pool is in the low thousands. Competing demand from a small number of high-margin firms (top 5 NYC firms, plus comparable London and Chicago firms) produces equilibrium compensation well above tech-company averages.
The dynamic is reinforced by performance-tied bonus structure. Quant firms typically allocate 30 to 60 percent of firm-wide P&L to compensation pools. As firm P&L per researcher has grown (with scale, capital deployed, and strategy diversification), per-researcher compensation has scaled proportionally. The largest quant firms have generated $100,000,000+ per researcher in some years; even at 30 percent allocation, this implies $30,000,000+ per researcher in compensation, with the actual distribution concentrated among senior partners.
The ML engineer who succeeds in quant is typically a research-strong PhD with demonstrated ability to identify and operationalise predictive signals in financial data. The skill set overlaps with academic ML research but with three additional emphases: (a) robustness under regime change (financial data is non-stationary), (b) speed of iteration (signal half-lives are often weeks to months), and (c) ability to work productively under tight confidentiality constraints (no external publication; limited discussion outside the firm).
For ML engineers from a tech background considering quant, the realistic entry path is typically through a smaller systematic fund or a bank systematic trading desk rather than direct entry to a top firm. The transition requires demonstrating ability to operate under the financial-data-specific constraints; once demonstrated, mobility upward into top firms is possible.
section qt-3 : risk profile vs frontier labs
Quant trading firms compete with top frontier AI labs for the same senior research talent, and the offer comparison is increasingly common. The two compensation structures are structurally different in ways that matter for long-run wealth-building.
Quant compensation is cash. Annual bonus, paid in cash within 12 to 18 months of the performance year, can be diversified immediately into broad market index funds or other long-term wealth-building assets. The cash structure also means the employee bears no single-employer concentration risk after the bonus is paid; if the firm fails or the employee leaves, the already-paid bonus is fully retained. The downside is no long-tail equity upside: if the firm grows ten-fold over the next decade, the employee's compensation grows roughly proportionally to their performance contribution rather than to firm equity value.
Frontier-lab compensation is illiquid pre-IPO equity. Realisation depends on the lab's eventual IPO or acquisition at a defined valuation. The expected value at current paper valuations is high but the variance is enormous: an unsuccessful frontier lab can produce zero realised equity value; a successful one can produce 10x to 100x the paper-value-at-grant. The concentration risk is severe; an ML engineer with $5,000,000 of unvested pre-IPO equity at a single frontier lab is structurally exposed to that lab's continued success.
For most senior ML engineers, the realistic conclusion is that quant offers superior near-term wealth-building (predictable, diversifiable, immediately liquid) while frontier-lab offers superior long-run optionality contingent on lab success. The choice often reflects the engineer's view of their own risk preference and their probability estimate of the specific lab succeeding.
section qt-4 : common questions
What is the average ML engineer salary at a quantitative trading firm?
Senior ML engineer total compensation at top-tier quant trading firms in 2026 is approximately $950,000 to $1,900,000 in a good performance year, comprising base salary $250,000 to $400,000 and performance-tied cash bonus $700,000 to $1,500,000 or more. The bonus is tied to firm-wide profit and loss; in a bad year the bonus can be substantially lower or zero. Second-tier and third-tier quant firms compensate below this but still above most tech-company equivalents.
Why are quant trading firm salaries so high?
Quant trading firms generate revenue per employee well above tech companies (often by an order of magnitude), so the marginal-revenue-product of a strong ML researcher is materially higher. The talent pool is narrow: top firms typically hire only PhD candidates from a small set of universities with research backgrounds in mathematical statistics, optimisation, machine learning theory, or high-energy physics. Strong supply constraint against very high demand pushes equilibrium compensation up. Performance-tied bonuses also align compensation with realised firm P&L, so high comp years coincide with high revenue years.
Do you need a PhD to work at a quant trading firm?
At the top tier, yes for most research roles, with rare exceptions. Top firms hire predominantly from PhD or post-doc backgrounds in fields with strong mathematical-rigour signalling. Research-strong undergraduates with strong competition placements (Putnam, IMO, ICPC) are sometimes hired directly into junior research roles. At the second tier and below, the PhD requirement is less strict. For engineering roles (versus research roles), a strong industry track record can substitute for the PhD; however, the engineering compensation at top firms is materially below the research compensation.
How do quant bonuses actually work?
Bonus is typically calculated as a share of firm-wide or desk-level P&L attributable to the employee's research contributions. The bonus pool at top firms can reach 30 to 60 percent of profit, distributed across research, engineering, and operations staff. Individual bonus depends on the firm's allocation methodology and the employee's measured contribution (which is itself an active research area at most quant firms). Bonus is paid in cash, typically in Q1 of the following year, with deferred portions in some firms. There is no equity grant analogous to tech-company RSUs.
What is the partner track at a quant firm and how much does it pay?
Most top quant firms operate a partner or principal track for long-tenured senior researchers with consistent multi-year P&L contribution. Partners receive a substantially larger share of firm profit (often through a separate vehicle: partner units, principal compensation pool, or named partner draws). Senior partners at the largest firms reportedly earn $10,000,000 to $100,000,000+ in good years, though this is highly variable and concentrated in a small number of individuals. The path from senior research engineer to partner typically takes 5 to 10 years of sustained P&L contribution. The selection is competitive: most senior researchers do not become partners.
How does quant comp compare to frontier-lab comp?
At the L5 senior level, top quant firms pay roughly comparable total compensation to top frontier AI labs in a good performance year. Top quant ($950k-$1.9M+) versus top frontier lab ($700k-$1.1M) shows quant winning on headline numbers, but with substantially different risk profiles. Quant comp is cash, paid annually, can be diversified immediately. Frontier-lab comp is illiquid pre-IPO equity, contingent on the lab's continued success. For ML engineers comparing offers, quant offers stronger near-term financial security; frontier-lab offers larger long-run optionality contingent on the lab's eventual valuation.
Are there quant ML jobs outside NYC?
Yes but the market is smaller. Citadel has a major Chicago presence. Two Sigma and DE Shaw operate Boston and Silicon Valley research centres. London hosts MAN AHL, Aspect Capital, GAM Systematic, and other large systematic funds. Hong Kong hosts a sizable Asian-quant cluster. Singapore is growing as a regional hub. For US-based quant ML engineers, NYC remains the dominant market, with Chicago a meaningful second; the smaller hubs typically employ 20 to 100 research engineers per location rather than the 100 to 500+ scale of the top NYC firms.
NYC metro
Quant capital, headquarter density
Frontier-lab comparison
Cash vs illiquid equity trade-off
PhD ML engineer salary
PhD requirement at quant firms
New York state
State + city tax math on quant bonuses
Total comp breakdown
Cash bonus vs RSU vs PPU mechanics
Negotiation playbook
Quant-vs-frontier-lab offer comparison